Find me at: | Linkedin| Github |Twitter | Udacity |
I am a Physicist. Currently, my interest is in Data Science, Machine Learning and Artificial Intelligence. I have two parallel thoughts for model building perspective: simulation of a mathematical model and optimization of the proposed model over a large data set. I think more in first perspective 'simulation of a mathematical model' when we are familiar with dynamics of the system under study and explore more in that part where analytical solution does not exist. Here we technically go from model to data generation. The second part 'optimization of the proposed model' comes with machine learning which goes from data to model building (opposite to the first one). Proper feature engineering and tuning of the model parameters are the most important part of the optimization.
Find my projects: | Artificial Intelligence | Self Driving Car |Machine Learning | Deep Learning|
Find my research: | Quantum Field Theory| Nature Inspired Computing|
Teaching: | Machine Learning for Aspirants | Computational Physics | Algorithms|
Project 1. Advanced Lane Line Detection Pipeline:
Built an advanced lane-finding algorithm using OpenCV for Hough Transforms and Canny edge detection, distortion correction, image rectification, color transforms, and gradient thresholding. Identified lane curvature and vehicle displacement. Detected highway lane lines on a video stream. Overcame environmental challenges such as shadows and pavement changes. Project Detail
Project 2. Vehicle Detection Pipeline :
Created a vehicle detection and tracking pipeline with OpenCV, histogram of oriented gradients (HOG), and support vector machines (SVM). Optimized and evaluated the model on video data from an automotive camera taken during highway driving. Project Detail
Project 3. Traffic Sign Classifier:
Built and trained a deep neural network to classify traffic signs, using TensorFlow. Experimented with different network architectures. Performed image pre-processing and validation to guard against overfitting. Project Detail
Project 4. Behavioral Cloning:
Built and trained a convolutional neural network for end-to-end driving in a simulator, using TensorFlow and Keras. Used optimization techniques such as regularization and drop out to generalize the network for driving on multiple tracks. Project Detail
Project 5. Sensor Fusion:
Implemented sensor fusion technique in C++ with an Extended Kalman Filter algorithm capable of tracking a pedestrian's motion in two dimensions and an Unscented Kalman Filter algorithm capable of accurately and performantly tracking a turning object. Project Detail
Project 6. Localization:
Implemented a 2-dimensional localization algorithm with Markove filter and particle filter in C++ capable of localizing a vehicle within desired accuracy and time. Project Detail
Project 7. PID Control:
Implemented a PID controller in C++ to maneuver a vehicle around a track. Project Detail
Project 8. Model Predictive Control:
Implemented Model Predictive Control to drive a vehicle around a track even with additional latency between commands. Project Detail